Final Drive, Lucky Me Part Two

[donsobeck]

I got my final drive in and I am back on the road and only $260 poorer.
I thought I would share this with you. When my first bearing went bad the balls wore out so you could feel and hear by rotating the rear wheel. So I would rotate the wheel after every ride to feel it and change the oil every 3000 miles more or less.
So with the new 17 ball bearing every was going great till I smelled gear oil and looked at the rear tire and it had oil on it(Oboy).
If the picture worked you can see that the cage for holding the balls let go and cut the oil seal, there was no noise or feeling that anything was wrong.
So you all can figure out how you are going to maintain your Fatal drive bearings.

DON

[dshealey]

Well, we now know the famous "17 ball" bearing fails in just the same way the older one did. No surprise to me. Same basic bearing construction, just a different manufacturer. I was just waiting for the failures to start on them too as the mileage built up on them.

[JPSpen]

Well there's your explanation. That sucker was made in France !!!!

Let's get a German bearing in there, or at least a Japaneese one. or if we just have to, an American one...Just kidding, a bit......

But now you've got me thinking..Has anyone tried toget this bearing/race
from someone else? Bearings and such are mostly standard sizes and I'm wondering if there is someone else who makes the same bearing..
I've got the numbers that are visible. 15-045X then the 5KF or maybe E. Were there any numbers on the bottom that I can't make out? I'm going to waste some time looking around about this.


Spence

[CWF]

How did you get the repair for only $260? You didn't replace that fine French bearing with Razor scooter bearings did ya?

[mneblett]

So with the new 17 ball bearing every was going great till I smelled gear oil and looked at the rear tire and it had oil on it(Oboy).

Don, it is very difficult to tell from the pic -- are you *sure* it's a 17 ball unit? It's not clear to me from the pic (counting the retainer "lumps" and remaining shards) that it's not a 19 ball unit.

[eljeffe]

That sure looks like a 19-ball bearing to me. As I recall, the 17-ball bearing looked a lot more beefier when I had mine replaced lasted year.

[mneblett]

That sure looks like a 19-ball bearing to me. As I recall, the 17-ball bearing looked a lot more beefier when I had mine replaced lasted year.
I had the same thought; moreover, the separtors look exactly like the one in my failed 19 ball unit.

[RaffyK]

That would be SKF a well known bearing manufacturer.

[Patric]

Hi ,I just used a ruler to measure around the bearing (using the picture ) in question .I set the depth gage on my pocket ruler the same as the spacing on the balls that were still where that belong . I counted 17 , but what do I know ...Patric ...

[donsobeck]

When I first had the bearing out I mark the outer race and counted the marks (DUD) then I counted the balls, there are 17 and the bearing didn't come apart.
If need be I could take more pictures.

DON

[mneblett]

When I first had the bearing out I mark the outer race and counted the marks (DUD) then I counted the balls, there are 17 and the bearing didn't come apart.
If need be I could take more pictures.

DON

That would be helpful, Don -- if you can get some more light in there, it would help resolve the confusion. I see what *may* be elongated and/or broken-off sections of bearing retainer, but I'm just not sure.

I just found this pic of a new 17 ball unit from the archives:
http://www.bmwlt.net/ubbthreads/downloadattachments.php?attach_id=7398?Cat= , and I have to say, the retainers look very similar. Maybe different retainers from different bearing vendors?? Either way, this isn't looking good for the 17 ball being an absolute cure :(

[KBandit]

in the absence of any official statement from BMW we are forced to assume that they have either:

1) taken steps to try to solve the rear drive problem, but failed
2) chosen to ignore the problem altogether, hoping it goes away

neither will do anything to help keep me as a loyal customer. the instant something better comes along i'm jumping ship.

[dshealey]

Well there's your explanation. That sucker was made in France !!!!

Let's get a German bearing in there, or at least a Japaneese one. or if we just have to, an American one...Just kidding, a bit......

Spence

The original "19 ball" bearing WAS a German bearing, by the bearing company FAG. The "new" bearing is by SKF, a Swedish based company, with manufacturing around the world.

I did a LOT of bearing research in the past on this, and did buy a bearing with a machined bronze retainer (by MRC), but totaled my bike before getting to try it out. Sold it to another member here, who has not installed it to my knowledge.

The basic bearing number, which can be cross referenced to any manufacturer is 61917 C3

[donsobeck]

I hope this works each hump is a ball.

DON

[DaveDragon]

Looks like 17 to me...

[Dochatley]

I got 17...

[jazzbass]

Don't know that the 17 ball bearing is going to be a perfect fix.

It looks like there is enough room for two more balls. Maybe two got
tossed when the bearing failed.

Does anybody know if that Manufacturer makes a 17 and a 19??

[KBandit]

It looks like there is enough room for two more balls. Maybe two got tossed when the bearing failed.

i was thinking (hoping) the same thing. i don't recall ever seeing a wheel bearing with that much empty space between the races. but i will readily admit to NOT being an expert on the subject by any stretch.

[jackd]

i was thinking (hoping) the same thing. i don't recall ever seeing a wheel bearing with that much empty space between the races. but i will readily admit to NOT being an expert on the subject by any stretch.

I'm no expert either but if the overall dimensions (outer size and inner size) must be the same and they have 2 less balls they have to have more air space.. between the balls.

[mneblett]

It looks like there is enough room for two more balls. Maybe two got tossed when the bearing failed.

Thanks, Don -- that's much better!

That's definitely a 17 ball bearing. :( If you use the spacing between the balls at the undamaged retainers, and walk that measurement around the races, 17 "undamaged intervals" fit neatly around the circle.

Crud. Our first "confirmed" 17 ball failure.

[dshealey]

The spacing between balls is a manufacturer's decision, based on how the retainer is going to be made, or just a cost decision in the cheaper bearings.

The machined bronze retainer bearing was a high quality industrial use bearing, and only had 16 balls, to make room for the more substantial bronze retainer.

The number of balls, contrary to what one would think, has very little to do with the bearing's load rating, because in radial loading only 2-3 balls will typically be taking the load at any one time regardless of the number.

[tliu]

...So you all can figure out how you are going to maintain your Fatal drive bearings.

DON

Thanks for sharing your experience. The picture of the bearing shows that the bearing cage got torn into pieces, and one visible nick on one of the bearing balls (which happened to be w/o cage and under light reflection.) The ball bearing cage looked pretty flimsy. Do you know if the cage was one single continous piece in a new bearing, and what material is it made of, pressed steel, brass, etc.? I assume that this is the new bearing BMW uses on the '03 MY and on. What's the difference between this one and the "old" bearing on '02 MY and prior?

You are right about checking the smoothness and quietness of rear wheel rotation, changing oil frequently, checking metal shaving at the drain plug, and DON'T ever let any dust get in the drive housing.

[motorman587]

I only have and counted two balls. :)

[RaffyK]

LOL...

http://www.central4wd.com/images/photos/Photo9920.8bullballs.jpg

[alindsay]

David,

I remember the picture of that bearing from the old site. It looked like a substantial piece. Do you recall where you got it and the part number? Too bad you didn't get to try it.

[dshealey]

This is the information I ordered mine by:

<<<<<<<<<<
THANKS FOR YOUR BEARING INQUIRY (85 X 120 X 18 WITH BRONZE RETAINER). WE CAN
OFFER YOU THIS SIZE IN THE MRC BRAND AT $ 144.25 EACH. DELIVERY WOULD BE 1
WEEK AFTER RECEIPT OF ORDER. IF YOU CAN USE THIS PART WITH A STEEL RETAINER,
THE COST WOULD BE $ 114.00 EACH.
WE ACCEPT ALL MAJOR CC, CAN SHIP C.O.D., OR YOU CAN OPEN AN ACCT WITH US.
CALL ME @ 800-225-4587 IF YOU NEED FURTHER HELP.

MARY ANN QUINTAL
EMERSON BEARING COMPANY

[tliu]

Anyone happen to have the part numbers for these two bearings made by FAG (old) and SKF (new)? Or better yet the tech specs.

[early1]

Thought these might help.
Some of David's pictures from the old site:
Old bearing
http://www.bmwlt.net/ubbthreads/downloadattachments.php?attach_id=4469?Cat=
http://www.bmwlt.net/ubbthreads/downloadattachments.php?attach_id=4468?Cat=
http://www.bmwlt.net/ubbthreads/downloadattachments.php?attach_id=4467?Cat=
http://www.bmwlt.net/ubbthreads/downloadattachments.php?attach_id=4466?Cat=
http://www.bmwlt.net/ubbthreads/downloadattachments.php?attach_id=4465?Cat=
http://www.bmwlt.net/ubbthreads/downloadattachments.php?
attach_id=4464?Cat=

New bearing
http://www.bmwlt.net/ubbthreads/downloadattachments.php?attach_id=4485?Cat=

[dshealey]

The specs are the same for both manufacturers.
Here are part numbers for a few manufacturers, but using 61917 C3 when talking to any bearing supplier will get you a bearing cross over, regardless of the manufacturer.

Koyo 6917 C3
FAG 61917 C3
SKF 61917 C3
NSK 6917 C3

Note that C3 is the designation for slightly looser than normal internal clearance measured in microns, may not be available easily from most manufacturers. BMW has specified that clearance, not sure how a normal fit bearing would react, but would not expect much difference, a toss up if it would be slightly better, or slightly worse.

[tliu]

The specs are the same for both manufacturers.
Here are part numbers for a few manufacturers, but using 61917 C3 when talking to any bearing supplier will get you a bearing cross over, regardless of the manufacturer.

Koyo 6917 C3
FAG 61917 C3
SKF 61917 C3
NSK 6917 C3

Note that C3 is the designation for slightly looser than normal internal clearance measured in microns, may not be available easily from most manufacturers. BMW has specified that clearance, not sure how a normal fit bearing would react, but would not expect much difference, a toss up if it would be slightly better, or slightly worse.

Dave, Thanks!

[Dman]

Note that C3 is the designation for slightly looser than normal internal clearance measured in microns, may not be available easily from most manufacturers. BMW has specified that clearance, not sure how a normal fit bearing would react, but would not expect much difference, a toss up if it would be slightly better, or slightly worse.

I wonder.....
Beings that the bearing is assembled with an axial preload that an increased clearance is called for to allow the ball track to be higher on the "shoulder" rather than at the bottom of the groove? Hmmm.............

Would think that the overall picture would be "slightly worse."

David, I still think that your idea of the machined brass retainer is the best route to take here.
(Especially after seeing a 17 ball bearing retainer pulled apart like the 19 ball ones!)

[JATownsend]

This is the information I ordered mine by:

<<<<<<<<<<
THANKS FOR YOUR BEARING INQUIRY (85 X 120 X 18 WITH BRONZE RETAINER). IF YOU CAN USE THIS PART WITH A STEEL RETAINER,
THE COST WOULD BE $ 114.00 EACH.

Hi David, your thoughts are always respected. For those of us with less of a knowledge base, might you offer some additional perspective?

What are the advantages using a bronze retainer?

Is it relevant to the 'Life' of the bearing or housing?

Why did you choose THIS bearing as a possible replacement to OEM?

If 'Fit' or installation is a (the) real issue, how does one ascertain that they may or may not be getting the proper install.

For example: After a rather strong rear-ending in Chicago last November, Chicago BMW replaced my original 19 ball unit with a newer 17 ball unit. What is the acid test for critical bearing install?

Thanks for your continued technical guidance on this board.

[motorman587]

Raffy, that was good, guess nobody else thought it was funny. Sorry, so serious here. :)

[gunny]

I see the problem, the bearing was made in France. French bearings in a German bike in the United States.

[dshealey]

My sole purpose in buying the machined bronze retainer bearing was this:

If the bearing fails, the retainer would not come apart like the steel strip retainers, wiping out the seal and letting the balls all shift toward one side. The retainer would keep the balls spaced properly, not letting the wheel wobble side to side, and since the seal would not get wiped out easily, oil would stay in. My thinking was it would make noise LONG before being non-rideable, allowing one to get to a good place to stop, maybe home or to a dealer. I did not think the bearing was going to last longer, just that a failure would likely be far less "sudden".

The reason I bought that particular bearing was that it was the ONLY one I could find with a bronze retainer. Looked hard too.

[JATownsend]

David, Thanks for the reply focusing on the retainer's ability to contain the damage a bit longer. That answers some questions.

Are these (Bronze) bearings any less likely to be affected by a poor installation process than the current crop of BMW bearings, or is that a 'Trick' question?

[tliu]

Dave,

Thanks for the insights. The best I can tell from MRC website, the bearings w/ dimension 85x120x18 are either 1916S (radial) or 1917R (angular contact.) Both bearings are full-complement (27 and 30 balls for 1916S and 1917R, resp'ly) w/o filling slots, i.e. no cages. They have an oil-filled speed rating of 6,000 rpm approx. Generally, bearing w/ filling slots, like the FAG/SKF bearings used on LT, are selected for higher-speed applications. Do you happen to know the max. rpm and static/dynamic loads requirements of LT rear drive bearings? Obviously, they are w/i the FAG/KSF bearing specs; but I have not been able to find them. Thank you in advance.

[dshealey]

No, improper installation would still cause brinelling of the balls/races, which shows up later as spalling. The bronze retainer does not relieve any pressure between the races and balls, nor can it protect them from overpressure if installed improperly, just performs much better as a ball "spacer", and is not distorted by the balls trying to overtake or overrun each other with cyclic loading. When this happens with the stamped/riveted strip steel retainers, the cyclic loading weakens and finally breaks the strip steel or the rivets. When this happens, as we know, a pretty fast failure ensues.

[dshealey]

Dave,

Thanks for the insights. The best I can tell from MRC website, the bearings w/ dimension 85x120x18 are either 1916S (radial) or 1917R (angular contact.)
Both bearings are full-complement (27 and 30 balls for 1916S and 1917R, resp'ly)
Neither of these could be used in the final drive. A full complement bearing is one with complete fill of balls, with no retainer/cage. These are only good for oscillatory use, where the rotary motion is oscillating back and forth, or for very slow rotary operation. Since the balls have no seperator, the surface of each ball is seeing rolling motion at the races, but will be driving it's leading side down against the trailing side of the ball in front of it which will be moving up. This scrubbing action of the ball surfaces will destroy it pretty quickly. Picture two tires rolling forward, then let the trailing one catch up to the leading one until the treads touch. No more rolling! That is what happens when a open wheel race car touches tires with another car, one is launched hard into the air. The retainer/cage in the bearings keeps the highly finished surfaces of the balls from contacting each other, and is softer than the balls so will not cause the ball surfaces to be damaged.

Angular contact bearings are only used in pairs, with opposing installation so one bearing takes thrust load in one direction, the other one the opposite direction. These are used in high thrust, lower radial load applications, and can be pre-loaded to give very high stiffness in all directions. The idea would be a good one in a final drive, but no room for the second bearing in the BMW design. There is a small preload in our drives between the large bearing and the quite small tapered roller bearing to the outside, but only to take the loads of the crown/pinion gears, not to take side loading of the wheel.

w/o filling slots, i.e. no cages. They have an oil-filled speed rating of 6,000 rpm approx. Generally, bearing w/ filling slots, like the FAG/SKF bearings used on LT, are selected for higher-speed applications.

You have something confused here, and I think reversed. Filling slots and cages are two unrelated things. The bearings on the LT have NO filling slots. Filling slots are used on high ball count bearings, such as full compliment bearings, because the way the balls are put in is to push the inner race one side until it contacts the inside of the outer race, then the balls are inserted into the wider opening made. As you can easily imagine, only a certain number of balls can be put in before the distance between the races prevents any more from being put in. Then the balls are pushed around and the races centered, then the retainer/cage is put in. Sometimes the manufacturer actually presses the outer race into a slightly oblong shape to get another ball or two in. In deep groove bearings which will not see high thrust loading, a small radiused "slot" is ground into each race to allow more balls to be put in. If the bearing was to see high thrust loading though, the balls would start to contact the slots during operation, causing rapid damage to the balls.

Do you happen to know the max. rpm and static/dynamic loads requirements of LT rear drive bearings? Obviously, they are w/i the FAG/KSF bearing specs; but I have not been able to find them. Thank you in advance.

No, I do not have any idea what the engineers have determined are the loads seen in the LT application. Obviously, a little more than they expected though. :D

[alindsay]

Thanks Dave.